#CLAUDE.md の記事

CLAUDE.md Evaluation Checklist | Diagnose Effectiveness with 5 Metrics

Rust Development with Claude Code | Clippy Integration and CLAUDE.md Configuration

How to Use Claude Code Rules | Differences and Use Cases vs. CLAUDE.md

A complete guide to Claude Code's rules feature (.claude/rules/): how it differs from CLAUDE.md, splitting rules into topic-based files, path-specific rules via paths frontmatter, user-level rules at ~/.claude/rules/, sharing via symlinks, writing rules that don't bloat, and verifying them with /memory.

How to Write CLAUDE.md | Memory Configuration and File Placement for Claude Code

CLAUDE.md is the instruction file that Claude Code reads at the start of every session. This article covers the difference between CLAUDE.md and automatic memory, the four scopes and their file paths, how to write effective instructions, @import, .claude/rules, AGENTS.md, /init, and troubleshooting — all organized based on the official specification.

The Ultimate CLAUDE.md Design | 7 Principles That Work for AI and the Bloat Trap

A breakdown of 7 design principles for an effective CLAUDE.md and the bloat traps to avoid. Claude Code auto-loads it at the start of every session, so its power is determined not by line count but by signal density. The Gotchas section offers the highest ROI, and cramming in style rules buries important directives under noise.

What Is Claude Rules MD | Designing AI Control with CLAUDE.md

A guide to writing CLAUDE.md, the file Claude Code automatically loads. Covers the four essential elements — role, code conventions, guardrails, and project knowledge — along with scope control via the paths field and @path syntax, differences from Cursor .mdc and AGENTS.md, and conversion using cursor2claude.

What Is CLAUDE.md, Now Adopted by Apple | The New Standard for AI-Driven Development

A guide to the role and implementation of CLAUDE.md, as seen in the Apple Support app repository. Covers the 5 essential elements to include—project overview, coding conventions, directory structure, common commands, and prohibited actions—plus how to leverage @claude in PR reviews as the new standard for AI-driven development.